DARPA Plots Supercomputing Future

The Omnipresent High Performance Computing program is seeking proposals to dramatically advance the performance and capabilities of future computing systems and enable ExtremeScale computing.

Elizabeth Montalbano, Contributor

June 30, 2010

2 Min Read
InformationWeek logo in a gray background | InformationWeek

Image Gallery: Government's 10 Most Powerful Supercomputers

(click for larger image and for full photo gallery)

Technology to support the development of the next generation of supercomputers is the goal of a new program launched by the military's research arm.

The Defense Advanced Research Projects Agency (DARPA) is seeking proposals for revolutionary approaches to "dramatically advance the performance and capabilities of future computing systems and enable ExtremeScale computing" as part of a new Omnipresent High Performance Computing (OHPC) program, according to a Broad Agency Announcement (BAA) available on FedBizOpps.Gov

The technology developed through the solicitation will support DARPA's Ubiquitous High Performance Computing (UHPC) program, which aims to develop a prototype for next-generation UHPC systems by 2018.

DARPA cited increasing demand for greater performance, higher energy efficiency, ease of programmability, system dependability and security as reasons for launching the program.

The agency said that current approaches to supercomputing R&D - which take an evolutionary approach rather than try to move the technology drastically forward - are inadequate.

As it considers proposals for the OHPC program, it will only consider those that offer what it considers revolutionary research ideas, according to the BAA. "Evolutionary research will not be selectable," the agency said.

Technologies DARPA is seeking proposals on include the following:

--How to reduce or manage power requirements for HPC systems, including memory and storage hierarchy;

--New hardware, software and language design for highly programmable systems that reduce the need for users to be aware of system complexity;

--Hardware and software that can improve system dependability, manage the component failure rate, and make security compromises such as sharing information and responsibility across a system's OS, runtime system, and applications; and

--Self-aware system software that can perform its own resource management, among other activities.

Supercomputers are necessary for complex algorithmic-based processes - such as 3D modeling -- that demand powerful and highly scalable processors.

A recent example of their usefulness is work the National Oceanic and Atmospheric Administration (NOAA) has been doing with researchers at several U.S. universities to create 3D models to map and predict the trajectory of the Deepwater Horizon oil spill in the Gulf of Mexico.

Firms have until Aug. 6 at 12:00 pm Eastern Daylight Time to respond to DARPA's OHPC BAA.

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like


More Insights